and
Mesopotamia in the second half of the third millennium BC. The 1964 discovery at the
Tell Mardikh site in Northern Syria of an ancient city from the second half of the third millennium BC completely altered archaeological knowledge of the time, as it indicated the existence of a contemporary urban culture during the
Early Dynastic Period of
Mesopotamia, within a geographic zone where, at the time, previous excavations had revealed nothing on the same scale. In agreement with
Ignace Gelb's theories on the subject of all inhabited centers in Syria of the same era, it appeared that the Tell Mardikh civilization's cultural identity did not necessarily fall within the
Semitic family. However, in 1968, the discovery at the same site of a statue bearing an ancient
Akkadian inscription, mentioning the king
Ibbit-Lim of
Ebla, soon contradicted this hypothesis. It therefore became possible not only to identify this city as the ancient city of Ebla, referred to in numerous Mesopotamian and Egyptian sources, but additionally, considering the strong linguistic connotations of the king's name, to specify the identity as
Amorite. It became necessary, however, to revise these conclusions again, after the 1974 discovery in the ancient ruins of a
Bronze Age palace (2400–2225 BC) of 42
cuneiform tablets, then of 17,000 others the following year, revealing a language different from
Amorite, which exhibited archaic morphological characteristics present in Akkadian, with incontestable lexical similarities to
West Semitic languages such as
Hebrew or
Aramaic.
Excavations were directed by Professor
Paolo Matthiae and the inscriptions translated by
Giovanni Pettinato. This opposition between a West Semitic
lexicon and an Akkadian
morphology led to controversies surrounding the nature of this language. For P. Fronzaroli, the opposition suggested an Akkadian
dialect that had undergone a strong Western influence. On the other hand, Giovanni Garbini favored a more nuanced approach, drawing attention to the fragility of a comparison with Akkadian, and pointing out that there is no other contemporary model with which to draw comparisons. In his "Considerations on the Language of Ebla", he highlighted the artificial character of this opposition between morphology and lexicon and noted that "Akkadian differs from Western Semitic as we knew it hitherto because the latter was documented only on the phase following Amorite innovation. If it is traced back to the time before these innovations, a northwestern pre-Amorite Semitic begins to emerge, which is concordant with Akkadian just because the latter preserved its earlier character after Amorite invasion". Essentially basing his study on the lexicon, G. Pettinato was nevertheless the first to announce in 1975 the discovery of a new Semitic language, to which he gave the name "Paleo-
Canaanite." Although the academic community was in favor of this idea, they were not unanimous regarding Pettinato's proposed name. In fact, while indicating advantageously its similarity to
Hebrew,
Ugaritic, or
Phoenician, the name proved nevertheless incapable of indicating its morphological roots in
East Semitic languages. G. Garbini then proposed the term "Paleo-
Syrian," but again, this proved just as inadequate to convey the Mesopotamian particularities and was not accepted. Therefore, without a name to fit this new language's different linguistic characteristics, "Eblaite" was finally chosen.
Nature of the documents Of the Eblaite corpus, whose publication began in 1974 as stated above, the majority of discovered documents are administrative or economic in nature, along with about a hundred historical tablets as well as some scholastic writings:
lexicons,
syllabaries, or bilingual texts. To this list, we must also add a few rare literary texts: fragments of
myths,
epics,
hymns,
proverbs, as well as some documents for
conjuration. From a linguistic perspective, although a great number of these documents were effectively written in
Sumerian, a rather large portion of these only used the language
ideogrammatically, as confirmed by certain
Semitic elements added to the
Sumerograms – such as morphological
markers, suffix pronouns, or certain prepositions – which reveal an underlying language distinct from Sumerian.Such writing practices obviously made approaching Eblaite difficult. Fortunately, some rare documents, bilingual letters or tablets, mostly written syllabically, enabled the breaking down of this graphical barrier and the clarification of our knowledge of this language. Of course, even if we add to this collection the
onomastic material, which in Semitic languages typically consists of short sentences, the portion of the Eblaite corpus that is usable from a
linguistic perspective remains relatively narrow and limited from a
morphological,
syntactical, or
lexical point of view.
The graphical barrier and writing practices The main difficulty faced by those studying the language of Ebla arose largely from issues in the
writing system. Indeed, Eblaite shares its
cuneiform writing system with the
Sumerian,
Akkadian,
Hittite,
Hurrian, and
Elamite languages, a graphical system where each symbol may have collectively or separately an
ideogrammatic and/or
phonetic value. In the first case, the symbol or chain of symbols simply signifies an idea that is understandable by way of its Sumerian meaning; in the second case, the symbol indicates, with a more-or-less large approximation based on writing practices, the form of an Eblaite term following a principle of syllabic decomposition. The comparative study of Eblaite symbols reveals some differences with the systems used by other schools of
scribes. On the other hand, the Eblaite syllabary, without being identical, bears significant similarities with that of the ancient Akkadian used in
Kish during the
Early Dynastic Period (DA II). In fact, three transcription practices appear in the Ebla texts: one exclusively syllabic, another using both syllabism and ideography, and the last largely employing the ideographic principle. Included in the first category is mostly the incantatory texts and the writing of
anthroponyms; in the second, the epistolary, historical, and literary documents, not to mention some diplomatic texts; and in the third, economic and administrative texts, relating to the management and stewardship of the palace where
ideography is a sufficient system for the writing of
realia. Qualitatively and quantitatively, this situation entirely resembles that of the
Mesopotamian corpus. Only a small portion of documents found are syllabic, compared to the large quantity of texts written using
Sumerian logograms. This led G. Pettinato to consider, at first, that these documents were written in Sumerian. Such a hypothesis obviously no longer holds today with regard to our understanding of the writing and formulation practices particular to Sumerian and Eblaite scribes. These graphical conventions are so specific that they are very often sufficient to identify the language underlying the ideograms. Thus, for example, the Sumerian practice of writing filiation following the formula X DUMU Y ("X son of Y") stands out from the Akkadian and Eblaite practice which prefers the phrasing X DUMU.NITA Y. However, if, as we just saw, we can identify a
signified of Semitic origin beneath a Sumerogram, it remains difficult to extract its signifier. Fortunately, the restoration of phonetic values to these symbols has been made possible by the existence of bilingual lexical lists, where each Sumerian ideogram has its Eblaite form specified in a glossary using syllabic writing. Even when the phonetic value of the word is specified, a whole series of
semantic problems remains, still obstructing our understanding. For example, when an Eblaite scribe uses the symbol
LUGAL meaning "king" in Sumerian, he transcribes it with its Akkadian value
šarrum but translates it as "dignitary." This simple example shows the gaps in interpretation that may result from reading Eblaite symbols while only considering their Sumerian values. As for the strictly syllabic system of writing, it is not free of issues either. The rarity of Vowel + Consonant -type symbols (VC) require certain approximations in the transcription of words. Thus we find the term
ʾummum "mother" syllabically rendered as
u3-mu-mu. Additionally, while Sumerian sometimes proceeds morphologically by reduplication of a word to make it plural, Eblaite reuses this practice with the same meaning, but transforming it into a simple graphical signified. In this way we find forms along the lines of nasi11-nasi11 to write the plural of nas11 "the people." Furthermore it is not uncommon that the writing presents a defective character, where all the morphological markers are not indicated:
ḫa-za-an šu-ba-ti = *
ḫazānum yimḫur "the mayor takes it." To these issues we can also add those connected with the intrinsic limits of the Sumerian writing system, incapable of rendering a portion of Semitic languages'
phonological system. As Diakonoff specifies, the Sumerian system is organized upon a tense~lax opposition and can only with great difficulty render the voiced~unvoiced opposition as well as the
emphatics of Semitic languages. Thus we find the syllables /da/, /ṭa/, and /ta/ transcribed with the same symbol DA, as well as the syllables /gu/, /ku/, and /qu/ with the same symbol GU. For the same reasons, it is equally impossible for the Sumerian writing system to render the
laryngeals and
pharyngeals of Eblaite. However, to overcome these difficulties, they used – just like ancient Akkadian – graphical conventions such as the use of the symbols E and MA to render the
phonemes /ḥ/ or /ʿ/, or else by playing on syllabic symbols which end in the vowel /e/, which is nothing but the
vocalic trace of one of the two preceding articulations. Additionally, as shown by the written forms
la-ḫa for /
laḫān/ or
ba-da-a for /
baytay/ for example, the phonemes /w/, /y/, /m/, and /n/ are not rendered graphically in the final or initial position. Taking these two examples again, notice that, for one, the quantity of the vowels is not rendered by the writing (the form
da-za-a for /
taṣṣaʾā/ "they will go out" shows us that double consonants face the same fate) and secondly, that the vowel /a/ is used equally to represent the syllables /ʾa/, /ya/, and /ay/. == Phonological system ==